Camelia, the Perl 6 bug

IRC log for #parrot, 2008-05-08

Parrot | source cross referenced

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 NotFound /* TODO warn or fill if there are holes */ ---> Maybe this is related to the problem?
00:11 particle[ventus] hrmm, could be....
00:22 pmichaud "svn status" is showing src/call_list.txt
00:22 pmichaud I'm not sure where that's coming from.
00:25 dalek r27390 | pmichaud++ | trunk:
00:25 dalek : [tools]:
00:25 dalek : * Update mk_language_shell.pl so that it doesn't attempt to configure
00:25 dalek :   shells created outside of the languages/ directory. (RT#53418)
00:25 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=27390
00:25 pmichaud 701 tickets.
00:26 wknight8111 you guys have been kicking butt today
00:26 wknight8111 (Parrot committers)++
00:32 Coke 700 now, looks like.
00:32 pmichaud merged two tickets.
00:32 Coke 58+642
00:32 purl 700
00:32 Coke that counts.
00:32 Coke chaos--
00:32 Coke clearly 700 was too easy of a goal.
00:34 pmichaud so, 650?
00:37 pmichaud I think 700 was a good initial goal.  I'd hate to have one we couldn't reach.
00:37 NotFound Inside parrot_mark_hash, if mark_key is true bucket->key must be non nulll, according the logic of the code. I put an assertion, and fails in the test case.
00:44 Coke nah, not going to actually suggest another goal. but am going to keep digging.
00:49 NotFound Dropping the assertion and not calling pobject_lives when null, the problem dissapear. Don't know if this is correct, however.
00:50 Tene Nobody's going to suggest 666 as the net goal?
00:55 Coke heh. I considered it then hushed up. =-)
01:04 kid51 joined #parrot
01:10 kid51 purl, is your messages service disabled?
01:10 Coke msg kid51 hi there.
01:10 purl Message for kid51 stored.
01:10 tetragon Is purl sleeping?
01:11 * tetragon grumbles about connection latency
01:12 kid51 Probably not, my wonderful Verizon DSL connection has dropped off twice in 15 minutes!
01:12 particl1 joined #parrot
01:13 kid51 For once, it's not a problem with the router's signal failing to reach my laptop.  It's the Internet connection itself.
01:14 pmichaud maybe the tubes are blocked.
01:14 pmichaud I blame Senator Ted Stevens
01:14 NotFound Looks like the problem is that delete_keyed_int sets the key in bucket to NULL, but mark_hash expects that it never be NULL.
01:15 NotFound s/mark_hash/parrot_mark_hash
01:23 japhb joined #parrot
01:59 dalek r27391 | pmichaud++ | trunk:
01:59 dalek : [pge]:
01:59 dalek : * Remove outdated todo from RT#44449.
01:59 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=27391
02:03 pmichaud 696.
02:21 Zaba joined #parrot
02:22 * particl1 looks for more tickets to add...
02:47 wknight8111 joined #parrot
03:01 janus joined #parrot
03:12 Psyche^ joined #parrot
03:30 Theory joined #parrot
04:00 tetragon joined #parrot
04:35 Psyche^ joined #parrot
04:44 Eevee joined #parrot
04:53 petdance joined #parrot
05:03 cotto_home #52976 has been resolved
05:04 cotto_home or at least worked around until the bug it exposed can be properly fixed
05:04 cotto_home the mmd code works, at any rate
05:16 cotto_home Is there any reason I shouldn't close that one myself?
06:12 uniejo joined #parrot
06:27 masak joined #parrot
06:29 masak joined #parrot
06:31 masak Jonathan: er, why did `make` in fact fail yesterday?
06:37 Infinoid cotto_home: unless you want to keep the ticket around to track the *proper* fix, feel free to close it (but be sure to post about the current situation)
06:40 allison joined #parrot
07:03 iblechbot joined #parrot
07:58 masak joined #parrot
10:10 desertmax joined #parrot
10:16 mire joined #parrot
10:45 dalek r27392 | fperrad++ | trunk:
10:45 dalek : [Lua] PGE
10:45 dalek : - changes the non-capturing subrule syntax from <?foo> to <.foo>
10:45 dalek : - uses the <foo: "bar"> syntax
10:45 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=27392
10:58 Zaba joined #parrot
11:31 Zaba joined #parrot
12:06 pmichaud cotto_home: (RT#52976)   if the issue no longer exists or can't be reproduced in the latest version, feel free to close the ticket.
12:12 tetragon joined #parrot
13:06 gryphon joined #parrot
13:09 Coke 14+683
13:09 purl 697
13:16 iblechbot joined #parrot
13:40 parrot-poke joined #parrot
14:17 particle msg notfound you forgot the attachment to #53850
14:17 purl Message for notfound stored.
14:22 ask_ joined #parrot
14:30 rdice joined #parrot
14:34 NotFound joined #parrot
14:34 NotFound Hello.
14:35 Coke goodbye!
14:35 * purl breaks out into "So Long, Farewall" from "The Sound of Music"
14:35 * Coke breaks purl's trachea.
14:36 * particle kicks purl in the tail
14:37 * Coke fight clubs. "I want you to hit me. As hard as you can."
14:37 NotFound How can I read purl messages?
14:38 Coke purl, messages?
14:38 purl To access purl's messages, msg me with the word "messages".
14:39 NotFound 'me' literally?
14:39 particle /msg purl messages
14:39 NotFound Ah, I was trying /msg msg
14:41 particle no problem, newbie ;)
14:42 NotFound Thanks.
14:45 NotFound particle: attached now, thanks.
14:47 NotFound particle: Have you seen #53890 ?
14:49 * particle sighs because he can't unbreak plumhead quickly :(
14:49 particle NotFound: lemme look
14:51 particle the patch looks good at first glance
14:51 particle lemme take a look at hash.c and make sure nothing's funny, and i'll apply
14:54 NotFound I wonder why only the key is nulled, and not the value.
14:55 masak Jonathan: `make` still doesn't build, just as yesterday. did the problem really go away, or did I misunderstand something?
14:56 Coke masak: perl6 or parrot?
14:57 masak Coke: perl6
14:57 masak on Mac OS X
14:58 moritz (whoever fixed make -j)++
14:58 Coke are you dong "make perl6" in the root, or cd'ing to languages/perl6 && make?
14:58 masak the latter
14:58 purl well, the latter is better
14:58 masak purl: I guess you'd know
14:58 purl masak: what?
14:58 masak nvm
15:05 Coke masak: works here, osx/intel
15:05 Coke this with a straight "perl Configure.pl" in parrot.
15:05 Coke do you have a ticket open or can you nopaste the error you're getting?
15:06 Coke http://paste2.org/p/26174 ?
15:07 Coke what version of osx/processer/compiler/configure flags ?
15:09 masak Coke: yes, that nopaste
15:10 masak Coke: OS X 10.4.11
15:10 masak I can nopaste the configure flags
15:10 pmichaud "unable to parse regex" makes it sound as though a PGE library needs rebuilding
15:10 pmichaud (given all of the PGE changes made yesterday)
15:11 Coke when you re-upped, did you rebuild parrot?
15:11 Coke a simple make *should* work, of course, but you might have to reconfigure or even realclean depending on bugs.
15:18 cosimo joined #parrot
15:28 mire joined #parrot
15:29 Tene Heh.  "depending on bugs" is an entertaining statement.
15:33 Coke ... do I amuse you?
15:33 Coke </goodfellas>
15:34 NotFound "Depending on pugs" may be even better.
15:42 masak I ran `make realclean`
15:42 particl1 joined #parrot
15:43 masak it works now
15:43 Tene masak: hopefully you ran 'make' after, yes?  I'd hope parrot works better compiled.
15:43 pmichaud perhaps there's a build dependency missing somewhere.  I regularly do 'make realclean' so I don't see those very often.
15:43 masak yes, I ran make after
15:44 masak trial-and-error made me realise that I needed to do that
15:46 sjansen joined #parrot
15:46 cjfields joined #parrot
15:46 * Coke ponders setting up something to better test our build deps.
15:53 * Coke drops off the grid for a bit.
16:16 iblechbot joined #parrot
16:27 desertmax joined #parrot
16:41 Theory joined #parrot
16:46 clunker joined #parrot
16:54 IllvilJa joined #parrot
17:06 dalek r27393 | fperrad++ | trunk:
17:06 dalek : [Lua]
17:06 dalek : - skeleton of a bytecode translator
17:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=27393
17:36 ambs joined #parrot
18:15 Ivatar joined #parrot
18:23 AndyA joined #parrot
18:56 paco Hi: parrot: 64-bit XCOFF executable or object module not stripped <-> aix - gcc , parrot compiles without problems, but I have some problems making /pmc2c.pl --c subproxy.pmc, my Storable.pm is old ..
18:56 moritz should I submit patches to rakudo as tickets, or should I just send them perl6-compiler?
18:56 paco Now Im thinking in installing native compiler ..
18:58 particle moritz: submit to rakudobug
18:58 particle rakudobug?
18:58 purl i think rakudobug is mailto:rakudobug@perl.org
18:58 particle paco: can you update Storable?
18:59 particle paco: and i wonder what the native compiler would do with parrot there.
18:59 paco particle : I tried, but Storable is reclaiming aix compiler, with gcc I have errors ..
18:59 NotFound particle: Have you seen chromatic's comment to my patch?
18:59 particle when you find out, sending a patch for PLATFORMS would be wonderful
18:59 particle NotFound: yes, i've just read it
19:00 particle NotFound: do you think you can hunt that down, or are you at a loss?
19:00 particle the hash code is more than a little bit nasty
19:00 moritz particle: too late, I was too impatient :/
19:00 NotFound I think that his suggestion implies modifying the documented semantic of oredered hash :?
19:00 particle moritz: np
19:00 particle NotFound: yes, that's correct
19:01 paco particle: with the native copiler I can update Storable ..
19:01 particle fact is, it's behavior isn't "officially" documented, it's just commented upon in the source
19:01 NotFound particle: doing that does not risk to break a lot of things?
19:01 particle paco: great. see how that goes
19:01 particle NotFound: it's possible, but lemme see...
19:02 cjfields joined #parrot
19:02 NotFound My current machines are not so fast to test any change with all languages.
19:03 particle OrderedHash isn't used in many languages, possibly only m4
19:03 particle NotFound: do you have ack?
19:03 NotFound Ah, looks more reasonable, then.
19:03 NotFound ack?
19:03 purl ack is probably at http://search.cpan.org/dist/ack/ or at http://petdance.com/ack/ or awesome
19:03 particle ack is also a grep-like tool for working-copies
19:03 purl okay, particle.
19:04 particle sudo cpan App::Ack
19:04 particle then, from a realclean'd working copy, you can do a search like "ack -al OrderedHash"
19:04 particle that will list all the filenames of files containing "OrderedHash"
19:05 NotFound Interesting, using grep in the tree is a shame.
19:05 particle grep-- ack++
19:05 NotFound Installing.
19:05 purl installing is, like, one thing, using it is another. or in the 'famous last words' category
19:05 paco installing :)
19:06 particle (more ack converts)++
19:06 Eevee oh I love ack++
19:07 particle nice patch for that long sub, moritz++
19:07 NotFound Failed test.
19:07 particle force install
19:07 purl force install is a bit primitive, and 5.8.4 and 5.8.7 may have different files
19:07 moritz thanks particle ;)
19:07 moritz required nearly no intelligence
19:07 particle the test in the current version is busted if you're root
19:07 particle the current development release has that test fixed
19:07 desertmax joined #parrot
19:07 NotFound particle: your fault, then ;)
19:08 particle yeah, i forgot. i use svn head
19:08 particle svn co https://ack.googlecode.com/svn/trunk ack
19:08 NotFound For tell me sudo, i mean.
19:08 particle yeah, indeed.
19:09 particle the dev release has been on cpan for some time, i should poke andy to make it non-dev
19:10 NotFound Looks like perl6 uses it.
19:11 NotFound And pugs, also.
19:11 particle pugs is unmaintained
19:11 particle you're not searching a 'realclean'ed repo i suspect
19:11 moritz and even more so the parrot backend
19:11 particle yeah, the parrot backend is bitrotten
19:13 dcat joined #parrot
19:14 NotFound Several libraries, also. HTTP SDL Pg
19:14 particle yes, those need to be tested
19:37 NotFound Testing a first attempt.
19:44 ambs joined #parrot
19:55 Jonathan moritz++ # cleaning up my mess :-)
19:55 NotFound Good, now two orderedhash tests segfaults.
19:55 moritz did anybody apply the patch?
19:56 Jonathan moritz: I'm currently packing my appartment up into boxes, so I'm kinda tied up, but it looks good to me; if nobody beats me to it, I'll certainly apply it when I get chance.
19:57 moritz Jonathan: ok
19:58 particle moritz et al: i've looked at a few good patches recently, but haven't the time to apply them atm
19:58 NotFound Can we make a lottery for precedence?
19:59 desertmax_ joined #parrot
20:06 wknight8111 joined #parrot
20:06 barney joined #parrot
20:07 * moritz tries to implement my $var1, $var2; in rakudo
20:09 NotFound particle: looks like chromatic has taken a different approach.
20:10 particle yeah. i'm not sure changing the semantics of OrderedHash should prevent us from fixing it the other way
20:11 particle but, it probably shouldn't be our first solution, unless it makes those semantics better match *Array and Hash PMCs.
20:11 particle so, does it?
20:11 Eevee moritz: no parens required now?
20:14 moritz Eevee: where?
20:14 Eevee my with multiple variables
20:14 NotFound particle: Don't know, lack experience.
20:14 moritz Eevee: no. Generally not required any more for lists
20:15 radhios joined #parrot
20:15 Eevee nice
20:15 moritz Eevee: just as you can say 'my @a = 1, 3, 4;' in Perl 6 (and it means the right thing)
20:16 Eevee (wart removal)++
20:16 particle NotFound: it's easy to test with a little pir
20:16 particle .sub 'test' :main
20:17 particle $P0 = new "Hash"
20:17 particle $P0['a'] = 1
20:17 particle delete $P0['a']
20:17 NotFound If you know what you want to test. I don't know what are the intended semantics.
20:18 particle what am i testing there... that there's no key of PMCNULL?
20:18 particle i expect in a Hash, there isn't.
20:18 particle i'm a bit distracted now, looking for an attorney
20:21 Andy particle: There's one bug I wanna fix before I push out 1.84
20:21 NotFound Testing chromatic's patch.
20:22 particle andy: anything i can help with?
20:22 paco I have a question, in a 64 bit platform, the default build must be 64 and passing --m32 we make a 32bits parrot ?
20:23 particle paco: supposedly, yes
20:23 paco ok
20:23 particle however, we should probably have an m=64 flag for completeness
20:23 paco yes, I think so
20:24 NotFound An m=8 for my Z80 machines will be nice ;)
20:24 particle urk.
20:24 paco NotFound: hahaha
20:27 NotFound chromatic's patch executes fine ./parrot --runcore=gcdebug t/pmc/orderedhash_9.pasm
20:28 particle excellent.
20:42 pmichaud YAPC::NA housing registration now available (http://conferences.mongueu​rs.net/yn2008/index.html)
20:44 particle can you refill the dining card?
20:52 moritz in future I should first test the things that I want to extend :/
20:52 pmichaud I dunno.  I just went with $40 for now and I can use cash for the rest.
20:55 particle yeah, i went with $20 for no good reason
21:12 * Tene takes note.
21:12 Tene yapcna hackathon before and after?
21:12 cjfields joined #parrot
21:12 particle parrot hackathon before
21:12 particle yapc hackathon after
21:13 Tene hm
21:13 Tene will there be parrot hacking at after?
21:14 particle wherever two or more of us are gathered...
21:14 Tene will there be two or more of us there after?
21:15 particle dunno. i'm leaving wed afternoon to go to PA via buffalo
21:15 allison I'm leaving thursday morning
21:15 pmichaud rumor has it that there will be some parrot hacking after, yes.
21:17 jjore I recall you probably don't actually want any of the food available with the dining card.
21:19 Tene pmichaud: know who will be there after?
21:20 particle jjore: there isn't much else around there to eat
21:22 allison IIRC you could also purchase food with cash
21:23 jjore Yes, that's also true. There was a late-night walkup place a few blocks away under the L which sold me a porkchop sandwich with mustard.
21:24 jjore They didn't do much of anything else and it wasn't awesome but it was ok and kind of interesting to get a porkchop on a bun as a fast food item.
21:24 jjore NW corner btw.
21:26 particle yay! my rebar has arrived.
21:26 allison whoops! My YAPC room payment confirmation page says "Conference: Your $1 order #$2" (spot the perl error!)
21:26 particle yep, i noticed that.
21:26 particle if you continue, things seem to work out okay.
21:27 NotFound says or prints?
21:27 particle it appears in the html
21:28 NotFound I mean, the \n
21:29 particle huh??
21:29 NotFound say or print .. "\n"
21:30 moritz what's your question?
21:30 NotFound A bad joke.
21:30 purl somebody said a bad joke was why don't people eat clowns.......because they taste funny. or What's brown and sticky?  A Stick! or what do you do with a dog that has no legs? take him for a drag. or what is brown and sounds like a bell? Dung.
21:30 NotFound At least purl understand me.
21:31 allison NotFound: I would assume prints, given the age of the donation system (predates say)
21:33 * particle considers adding an opp to mk_language_shell.pl
21:33 particle hrmm. instead, i'll nap, and maybe do it afterwards.
21:33 NotFound allison: surely they wait to perl6 to update it.
21:34 Tene so nobody knows of any parroters staying after for hackathon?
21:34 particle check with coke, jhorwitz, chromatic, etc
21:35 allison NotFound: well, it's in 5.10, I just doubt the donation system is running on 5.10
21:35 Tene you mean with people not here right this minute?  you mean plan for the future and follow through?
21:35 Tene Sounds dangerous.
21:35 moritz there isn't really much at all which is in 5.10 (only)
21:36 NotFound So they wait for perl6 before updating the program.
21:39 teknomunk joined #parrot
21:40 allison NotFound: ah, an optimist! :) I'm sure the program will be updated at some point...
21:43 NotFound allison: if someone understand my bad jokes in bad english, all is possible.
21:51 allison :)
22:00 wknight8111 What's the status with docs/book? I mean, it's obviously out of date, but is it intended to be updated?
22:01 moritz wknight8111: it would be great if somebody updated it. That's why it was made open source.
22:01 wknight8111 Good. I'm going to go through it like a hurricane :)
22:11 NotFound There in an inconsistence in class.pmc:add_attribute: the =item says: "Creates a new class if the current class has been instantiated." but in the code:/* If we've been instantiated already, not allowed. */
22:13 allison the code is correct
22:13 NotFound Always ;)
22:13 allison maybe not always, but in this case it is. The text reflects an earlier design that we abandoned.
22:14 NotFound The comment reflects what the next line does, yes.
22:19 Theory joined #parrot
22:22 pmichaud joined #parrot
22:22 Jonathan joined #parrot
22:22 wolverian joined #parrot
22:25 allison left #parrot
22:30 PerlJam joined #parrot
22:35 Juerd joined #parrot
22:35 leo joined #parrot
22:46 parrot-poke documentation is the best mentation
23:14 tetragon joined #parrot
23:14 particl2 joined #parrot
23:27 rdice joined #parrot
23:44 wolverian joined #parrot
23:46 cjfields joined #parrot
23:48 cjfields_ joined #parrot
23:52 Hinrik joined #parrot
23:52 Hinrik hi
23:53 Hinrik I was "Perl 6 and Parrot Essentials" and wondering, why would you choose to write pure PASM instead of PIR?
23:53 Hinrik +reading
23:57 Tene Hinrik: you wouldn't, really.

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

Parrot | source cross referenced